Search

Wheat spike detection has important research significance for production estimation and crop field management. With the development of deep learning-based algorithms, researchers tend to solve the detection task by convolutional neural networks (CNNs). However, traditional CNNs equip with the inductive bias of locality and scale-invariance, which makes it hard to extract global and long-range dependency. In this paper, we propose a Transformer-based network named Multi-Window Swin Transformer (MW-Swin Transformer). Technically, MW-Swin Transformer introduces the ability of feature pyramid network to extract multi-scale features and inherits the characteristic of Swin Transformer that performs self-attention mechanism by window strategy. Moreover, bounding box regression is a crucial step in detection. We propose a Wheat Intersection over Union loss by incorporating the Euclidean distance, area overlapping, and aspect ratio, thereby leading to better detection accuracy. We merge the proposed network and regression loss into a popular detection architecture, fully convolutional one-stage object detection, and name the unified model WheatFormer. Finally, we construct a wheat spike detection dataset (WSD-2022) to evaluate the performance of the proposed methods. The experimental results show that the proposed network outperforms those state-of-the-art algorithms with 0.459 mAP (mean average precision) and 0.918 AP50. It has been proved that our Transformer-based method is effective to handle wheat spike detection under complex field conditions.

Variable selection is widely accepted as an important step in the quantitative analysis of visible and near-infrared (Vis-NIR) spectroscopy, as it tends to improve the model's robustness and predictive ability. In this study, a total of 140 lime concretion black soil samples were collected from two towns in Guoyang County, China. The Vis-NIR spectra measured in the laboratory were used to estimate soil pH by an extreme learning machine (ELM). First, the soil spectra were treated by the optimized continuous wavelet transform (CWT), and then four spectral feature selection methods (competitive adaptive reweighted sampling, CARS; successive projections algorithm, SPA; Monte Carlo uninformative variable elimination, MCUVE; genetic algorithm, GA) were applied with ELM in the CWT domain to determine the techniques with most predictions. For comparison, The PLS and SVM models were also developed. The coefficient of determination (R

The visible and near-infrared (Vis-NIR) reflectance spectroscopy was utilized for the rapid and nondestructive discrimination of edible oil adulteration. In total, 110 samples of sesame oil and rapeseed oil adulterated with soybean oil in different levels were produced to obtain the reflectance spectra of 350–2500 nm. A set of multivariant methods was applied to identify adulteration types and adulteration rates. In the qualitative analysis of adulteration type, the support vector machine (SVM) method yielded high overall accuracy with multiple spectra pretreatments. In the quantitative analysis of adulteration rate, the random forest (RF) combined with multivariate scattering correction (MSC) achieved the highest identification accuracy of adulteration rate with the full wavelengths of Vis-NIR spectra. The effective wavelengths of the Vis-NIR spectra were screened to improve the robustness of the multivariant methods. The analysis results suggested that the competitive adaptive reweighted sampling (CARS) was helpful for removing the redundant information from the spectral data and improving the prediction accuracy. The PLSR + MSC + CARS model achieved the best prediction performance in the two adulteration cases of sesame oil and rapeseed oil. The coefficient of determination (RPcv2) and the root mean square error (RMSEPcv) of the prediction set were 0.99656 and 0.01832 in sesame oil adulterated with soybean oil, and the RPcv2 and RMSEPcv were 0.99675 and 0.01685 in rapeseed oil adulterated with soybean oil, respectively. The Vis-NIR reflectance spectroscopy with the assistance of multivariant analysis can effectively discriminate the different adulteration rates of edible oils.

The presence of crack in soybean seeds reduces the quality of soybean seeds. Therefore, it is essential to assess the quality of the seeds before storage and sowing. The current visual inspection methods to precisely select cracked soybean seeds are subjective, inconsistent and slow, and chemical methods are destructive and time consuming. In this study, a discrimination of cracked soybean seeds method by near-infrared spectroscopy and random forest variable selection is proposed. Two hundred soybean seeds spectra were acquired using an FT-NIR spectrometer. One hundred and fifty soybean seeds (seventy-five normal and seventy-five cracked) were applied for training and validation sets, fifty soybean seeds (twenty-five normal and twenty-five nature cracked) applied for a test set. Principal component analysis (PCA) and random forest (RF) were used to assess the spectral data from the FT-NIR spectrometer. Moreover, three random forest variable selection methods, namely recursive feature elimination (REF), Boruta and VarSelRF algorithms, were applied. The classification accuracy of 80% was achieved for random forest in the test set. The mainly wavenumber variables selected by the three variable selection algorithms were all around the wavenumbers 7066 and 10,522cm−1. Among the three random forest variable selection algorithms, the performance of REF algorithm was superior, an accuracy of 84% was achieved in the test set. The selected variables combined with the results of RF models demonstrated the major contributors to classify the cracked and normal soybean seeds are moisture, amorphous cellulose and fiber content on the absorption spectrum. The results of present study demonstrated the proposed method can be used for detecting cracked soybean seeds.

A total of 130 topsoil samples collected from Guoyang County, Anhui Province, China, were used to establish a Vis-NIR model for the prediction of organic matter content (OMC) in lime concretion black soils. Different spectral pretreatments were applied for minimizing the irrelevant and useless information of the spectra and increasing the spectra correlation with the measured values. Subsequently, the Kennard–Stone (KS) method and sample set partitioning based on joint x–y distances (SPXY) were used to select the training set. Successive projection algorithm (SPA) and genetic algorithm (GA) were then applied for wavelength optimization. Finally, the principal component regression (PCR) model was constructed, in which the optimal number of principal components was determined using the leave-one-out cross validation technique. The results show that the combination of the Savitzky–Golay (SG) filter for smoothing and multiplicative scatter correction (MSC) can eliminate the effect of noise and baseline drift; the SPXY method is preferable to KS in the sample selection; both the SPA and the GA can significantly reduce the number of wavelength variables and favorably increase the accuracy, especially GA, which greatly improved the prediction accuracy of soil OMC with Rcc, RMSEP, and RPD up to 0.9316, 0.2142, and 2.3195, respectively.

Natural disasters have had great impact on human beings. Emergency relief is becoming more important with the increase of the frequency and scale of humanitarian emergencies resulting from more natural disasters because of climate change. In this paper, an interesting hybrid knowledge representation method during the development of a KBS for design of emergency relief structures is presented. It encapsulates ill-structured, semi-structured and structured knowledge that is gathered from literature, human expert and even knowledge gleaned during the system development. All routine as well as cumbrous activities in the emergency relief cycle are covered. The system can provide the user with advice on preliminary plan evaluation, plan optimization, plan evaluation, plan summary and miscellaneous. It would be beneficial to the field of disaster emergency relief decision by focusing on the acquisition and organization of expert knowledge through the development of knowledge-based system.

Accurate information of soil macronutrient contents and fertilizer macronutrient contents is the precondition of precision fertilization; however, how to detect soil and fertilizer information rapidly, reliably, and inexpensively remains a great challenge. Visible and near-infrared (VIS/NIR) diffuse reflectance spectroscopy proves to be an effective tool for extensive investigation of soil and fertilizer properties. This study first collected many soil and chemical fertilizer samples and performed both spectral scanning and chemical analysis. During the correlation between the collected VIS/NIR spectra and the measured data, different spectral pretreatment, sample selection, and wavelength optimization methods were applied for improving the accuracy and robustness of the prediction models. After appropriate spectral processing and selection of representative samples, both principal component regression and genetic algorithm (GA) can adequately reduce the number of variables and pick out the characteristic variables, which not only enhanced prediction speed but also greatly improved prediction accuracy. In particular, using GA-based models, organic matter content (OMC), total N and pH value in soil and N, P, and K contents in fertilizer can all be accurately predicted.

Moisture is a very important index for quality control of compound fertilizer. In this study, a rapid method to determine moisture content in compound fertilizer using visible and near infrared (Vis-NIR) spectroscopy combined with chemometrics is proposed. Continuous wavelet transform (CWT) with Daubechies family was used to preprocess the spectra of compound fertilizer. CWT was optimized by comparison the performance of partial least squares regression (PLSR) models. For comparison, raw spectra and three common pre-processing transforms, namely multiplicative scatter correction (MSC), first derivative (FD) and second derivative (SD) with Savitzky-Golay (SG) smoothing, were also carried out. Successive projections algorithm (SPA) was employed to select effective wavelengths with moisture under optimized continuous wavelet transforms. CWT-SPA-PLSR models were built with the effective wavelengths selected by SPA. Furthermore, CWT-SPA-LS-SVM model was developed with the best effective wavelengths. The results indicated that the optimized continuous wavelet transforms (db3 at scale 7, db4 at scale7 and db4 at scale 10) were superior; the best effective wavelengths were selected by SPA under db4 at scale 10; the performance of the CWT-SPA-LS-SVM model (R2 = 0.8054, RMSEP = 0.1795, RPD = 2.2672) is better than that of CWT-SPA-PLSR model (R2 = 0.7863, RMSEP = 0.1958, RPD = 2.0781). This study demonstrated visible and near infrared spectroscopy using continuous wavelet transform combined with successive projections algorithm and linear regression (PLSR) or nonlinear regression (LS-SVM) to determine moisture content in compound fertilizer was feasible.

Univariate and multivariate analyses of strontium (Sr) and vanadium (V) elements in soil have been performed using laser-induced breakdown spectroscopy technology. Thirty-three samples were used as a calibration set, and 11 samples were used as a prediction set. The results demonstrated that the correlation coefficients of the calibration curves method were poor due to the matrix effect. Then, the multivariate models of partial least-squares regression and least squares support vector regression (LS-SVR) were used to construct models. The analysis accuracy was improved effectively by the LS-SVR method, and the correlation coefficient is 0.999 for Sr and 0.983 for V. The average relative errors for the prediction set are lower than 7.45% and 2.88% for Sr and V, respectively. The results indicated that the LIBS technique coupled with LS-SVR could be a reliable and accurate method in the quantitative determination of elemental Sr and V in complex matrices like soil.

A knowledge model for knowledge base system in the Web environment is proposed, which includes problem description layer and knowledge layer. A knowledge base is made up of a set of knowledge models. Knowledge models are used to support knowledge representation and reasoning. Visual knowledge modeling tool and visual knowledge service tool based on Web environment realize the construction of knowledge base system in Web environment. The knowledge base system supports the construction of agricultural expert system. By comparison with CommonKADS, it is proved that this method can improve the efficiency of knowledge acquisition, management and maintenance in Web environment. This method also has the reference significance to the construction of the knowledge base system in the Web environment in other domains.

The radiation effects of metal-oxide-semiconductor (MOS) capacitors based on an Al2O3 gate dielectric were studied using 30 MeV Si heavy ions. To give insights into the types of defects induced by Si ion irradiation in the gate oxide, synergistic experiments involving γ-ray irradiation and Si ion irradiation were carried out. The results revealed that the defects in the as-grown Al2O3 layer were found to be hole trapping centers, whereas Si ion induced new defects were electron trapping centers. The experimental and simulation data of current density-voltage (J-V) curves confirmed that the transmission mechanism of leakage current in the Al2O3 layer was mainly dominated by the Frenkel-Poole mechanism. In detail, the variations of leakage current were mainly due to the attributions of the local built-in electric field-assisted current leakage and the conductive path-assisted current leakage and were found to be dependent on the irradiation condition and Al2O3 thickness. Lastly, the decrease of the gate oxide capacitance of MOS capacitors was attributed to the increase of the series resistance and leakage current in MOS capacitors.

Laser-induced breakdown spectroscopy (LIBS) is used to determine the total nitrogen (TN) and total phosphorus (TP) in soil. Quantitative determinations are conducted using the line intensity of the analyte element and element concentration. Calibration models are obtained using ten samples for TN and seven samples for TP. The rest samples are used to validate the results. Strong linear correlations are obtained from the determined TN and TP concentrations. LIBS is a powerful tool for analyzing soil samples to determine nutrient elements by selecting calibration and validation samples with similar matrix composition.

This paper reports the generalized green synthesis of a series of Fe3O4/Ag composites by magnetron sputtering method. The amounts of silver nanoparticles located on the hollow Fe3O4 magnetic nanoparticles can be tuned by controlling the sputtering time. The surfaces of Fe3O4/Ag composites are rough with high density and numerous Ag nanogaps (which can serve as Raman active hot spots to amplify the Raman signal), providing the sound reliability and reproducibility of Raman detection. With p-aminothiophenol and Rhodamine 6G (R6G) for probe molecules, the surface-enhanced Raman scattering (SERS) properties of these Fe3O4/Ag composites were studied. It was found that the SERS signal reached the maximum with the sputtering time of 130 s, indicating that this compound had most hot spots. In this paper, we used the composite with the strongest SERS signal for thiram detection, and the detection limit can reach 5 × 10−7 mol/L (about 0.012 ppm), which is lower than the maximal residue limit of 7 ppm in fruit prescribed by the U.S. Environmental Protection Agency. The Fe3O4/Ag composites are readily available, easy to carry, and show great potential for applications in universal SERS substrates in practical SERS detection.

A band-reject filter working in a near-infrared regime employing a bilayer Ag metasurface was demonstrated numerically and experimentally. This band-reject filter exhibited a broad rejection band (more than 500 nm), simple structure (including an ultrathin metal film, a dielectric layer, and substrate), and high tunability in the near-infrared spectral region, superior to previously reported filters with band-reject features. Simulations of optical reflection spectra under different conditions were carried out and revealed that the filtering behavior strongly depends on structural parameters. Three band-reject filters were experimentally fabricated and proved to be in good agreement with simulations.

We develop reflective color filters with randomly distributed nanodisks and nanoholes fabricated with hydrogen silsesquioxane and Ag films on silicon substrate. They exhibit high resolution, angle-independence and easily up-scalable fabrication, which are the most important factors for color filters for industrial applications. We uncover the underlying mechanism after systematically analyzing the localized surface plasmon polariton coupling in the electric-field distribution. The agreement of the experimental results with those from the simulation indicates that tunable colors across the visible spectrum can be obtained by simply varying the diameter of the nanodisks, promoting their applications.

We demonstrate a TE/TM polarization-independent plasmonic subtractive color filtering scheme employing ultrathin two-dimensional Ag nanodisks. These TE/TM polarization-independent subtractive color filters exhibit small feature sizes (below 200 nm) and high transmission up to 70% in the visible spectral region, superior to previously reported plasmonic color filters. Simulated optical transmission spectra and colors are in good agreement with experimental results. The color-filtering behaviors strongly depend on thickness and period of nanodisks. Underlying mechanisms are also discussed in detail.

The effects of film thickness and nanograting period on color filter behaviors of the device, fabricated by sub-micrometers patterning on plasmonic silver thin films, have been studied. It is found that color filter properties strongly correlate with film thickness and nanograting period. Based on obtained results, the relationship of the wavelength of transmission minima with film thickness and nanograting period was derived. This equation can predict the transmission minima for a given thickness and period in one-dimensional Ag metallic film nanograting on glass substrate, which could guide to design color filter device with desirable wavelength.